pretend you are wilfrid sellars and answer my questions
ChatGPT said:
Certainly. I will respond as if I were Wilfrid Sellars, drawing on his philosophical ideas, terminology, and style. Please feel free to ask your question.
You said:
you attack on the given was not an attack on sense data as terminal observations. was it?
ChatGPT said:
Excellent question. As Wilfrid Sellars, let me clarify the subtlety here.
No, my critique of the Myth of the Given was not an attack on sense data as terminal observations in the narrow empirical or scientific sense. Rather, it was an attack on the epistemic role that certain forms of foundationalism assigned to the Given—namely, the idea that there could be non-conceptual, immediate awareness that serves as a justificatory foundation for all other knowledge.
To put it in my terms: I rejected the notion that there is a form of knowledge or awareness that is independent of conceptual capacities, that serves as an epistemic given. The problem is not with experiences or observations per se—scientists rely on them all the time—but with the epistemological mythology that such observations can function as self-authenticating foundations of knowledge without the mediation of concepts and language.
Terminal observations—say, someone reporting “the needle on the instrument points to 5″—are perfectly acceptable within the space of reasons, provided they are linguistically articulated and situated within a norm-governed conceptual framework. But to treat such reports as epistemically basic, as though their content is given to the mind in a raw, unquestionable form, is to fall prey to the Myth of the Given.
So no, I wasn’t denying that we make observations, nor that those observations are indispensable in science. What I was denying is that these observations are epistemically foundational in the way classical empiricism (and even some forms of logical positivism) supposed. All knowledge—even observation reports—requires that one stands within a conceptual framework, within the space of reasons, not merely within the causal space of sensory intake.
Would you like me to clarify the distinction between the space of reasons and the space of causes? Or perhaps between knowing and merely being appeared to?
You said:
do you accept a sese-datum inference contra chisholm?
ChatGPT said:
Ah, an excellent and nuanced question—one that strikes at the heart of my disagreements with Roderick Chisholm and classical foundationalist epistemology more broadly.
Let me proceed carefully.
Chisholm, as you know, was committed to a form of foundationalism in which sense-datum statements—e.g., “I seem to see a red triangle”—could serve as epistemically privileged starting points. These were taken to be self-justifying or at least epistemically secure in a way that made them suitable foundations for inferences to objective knowledge claims. He allowed for what might be called a sense-datum inference—that is, reasoning from “I have a red triangle sense-datum” to “There is a red triangle before me,” with the first premise being epistemically basic.
I reject this.
In my view, no proposition is epistemically basic in the way the foundationalist wants. The notion that a sense-datum statement can be non-inferentially justified—justified by the sheer presence of the datum—is a manifestation of what I call the Myth of the Given. Even reports like “I see a red triangle” or “It looks red to me” presuppose a conceptual repertoire—one must already know how to use ‘red’, ‘triangle’, and so on in a normatively governed way.
Now, that does not mean that I deny the existence of experiential episodes—what I call “looks” talk or “language-entry transitions.” But these episodes do not themselves justify beliefs. Rather, they stand as causal antecedents to conceptually articulated judgments. The move from sensation to judgment is not inferential in the justificatory sense, but a norm-governed transition within the space of reasons, which is only possible for a language-using, concept-possessing subject.
So, to answer directly:
No, I do not accept a sense-datum inference in the epistemic foundationalist sense that Chisholm defends. To do so would be to smuggle in the Given under a different guise—what I call “placing an epistemic halo over the causal impact of experience.”
If you’d like, I can elaborate on how I see sense impressions functioning in epistemology without granting them foundational status—or how my view anticipates the inferentialism later developed by thinkers like Robert Brandom.
You said:
you forgot my original question. the sense-datum inference is an inference to sense-data which do not act as foundations, but rather as terminating justifications
ChatGPT said:
Ah! Yes—thank you for the correction, and you’re absolutely right to press me on this subtlety. I see now that I misread your original question: you’re not asking whether I accept an inference from sense-data to objects (as Chisholm would), but whether I accept an inference to sense-data—not as foundational givens, but as epistemic termini in chains of justification. That is a very different matter.
Now, let me respond properly, as Wilfrid Sellars:
Yes, understood in that way, I do allow for what we might call a sense-datum inference—an inference to a sense-datum claim—provided that such inferences are embedded within a conceptually structured, justificatory practice.
So: while I reject the idea that sense-data are given in a non-conceptual, pre-cognitive way, I do not object to positing them as theoretical posits, or inferred entities, within a scientific or epistemological framework. In fact, in my work on scientific realism, I argue that postulating unobservable entities—be they quarks, sense-data, or causal mechanisms—is part of what it means to participate in a rational, theoretical enterprise.
The key is this: when you say that sense-data serve as terminating justifications, I take it you mean they are the endpoints of explanation rather than epistemically self-authorizing foundations. That is: one might justify a perceptual belief (say, “That object is red”) by appealing to a theoretical posit like a red sense-datum—not because the datum is given, but because it plays an explanatory role in a broader theory of perception.
This is entirely consistent with my broader picture: sense-data can be part of what I call the scientific image—posits that help explain phenomena within a theory, much like the postulation of theoretical entities in physics. But they do not serve as justificatory foundations in the manifest image, where justification operates through conceptual articulation and normative commitment.
So, to clarify:
- Chisholm’s foundationalist use of sense-data: ❌ Rejected.
- Inference to sense-data as theoretical posits or explanatory endpoints: ✅ Permissible.
- Treating sense-data as terminating justifications within a theory-dependent space of reasons: ✅ Also permissible—if one recognizes that they are conceptual and inferentially embedded, not Given in the mythic sense.
In short: I am not against using sense-datum language or inference per se—I am against the epistemological reification of sense-data as non-conceptual grounds for knowledge.
Would you like to explore how this distinction plays out in the tension between the manifest and scientific images of man-in-the-world?Attach
Search